Face Video Retrieval via Deep Learning of Binary Hash Representations
Authors: Zhen Dong, Su Jia, Tianfu Wu, Mingtao Pei
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our method achieves excellent performances on two challenging TV-Series datasets. The paper includes sections such as 'Experiments', 'Experimental Settings', and 'Results and Discussions', detailing empirical evaluations and comparisons. |
| Researcher Affiliation | Academia | Zhen Dong1 and Su Jia2 and Tianfu Wu3,4 and Mingtao Pei1 1. Beijing Laboratory of IIT, School of Computer Science, Beijing Institute of Technology, Beijing, China 2. Department of Applied Mathematics and Statistics, Stony Brook University, Stony Brook, USA 3. Institute of Sensing Technology, Beijing University of Posts and Telecommunications, Beijing, China 4. Department of Statistics, University of California, Los Angeles, USA |
| Pseudocode | Yes | Algorithm 1: Algorithm of low-rank discriminative binary hashing |
| Open Source Code | No | The paper does not provide an explicit statement or link for open-source code for the described methodology. |
| Open Datasets | Yes | The deep CNN is pre-trained on the Image Net dataset which has more than 1.2 million images of 1000 categories to obtain good initializations. We use the released Alex Net (Krizhevsky, Sutskever, and Hinton 2012) model which is trained on Image Net dataset for convenience. ... We use the ICT-TV dataset (Li et al. 2015c) to evaluate the proposed method. |
| Dataset Splits | No | The paper specifies training and testing sets, but does not provide details on a separate validation set or explicit split percentages for training, validation, and testing. |
| Hardware Specification | No | The paper does not specify any particular hardware components (e.g., GPU or CPU models, memory sizes) used for running experiments. |
| Software Dependencies | No | The paper does not specify software dependencies with version numbers (e.g., specific deep learning frameworks, libraries, or programming language versions). |
| Experiment Setup | Yes | The deep CNN is trained by the stochastic gradient descent with momentum of 0.9 and weight decay of 0.0005. The mini-batch size of the training samples is 128, and the triplets are randomly generated based on the labels. |